Journal Article

Cargo-cult statistics and scientific crisis

In this new article in Significance magazine, BIDS Senior Fellow Philip B. Stark and co-author Andrea Saltelli (an adjunct professor in the Centre for the Study of the Sciences and the Humanities at the University of Bergen) comment on the poor practices that result in the failure of scientific results to be reproducible and replicable, and offer an historical account of how these issues may have originated.  


Slingshot: cell lineage and pseudotime inference for single-cell transcriptomics

Single-cell transcriptomics allows researchers to investigate complex communities of heterogeneouscells. It can be applied to stem cells and their descendants in order to chart the progression from multipotentprogenitors to fully differentiated cells. While a variety of statistical and computational methods have been proposedfor inferring cell lineages, the problem of accurately characterizing multiple branching lineages remains difficultto solve.

The Types, Roles, and Practices of Documentation in Data Analytics Open Source Software Libraries; A Collaborative Ethnography of Documentation Work

Data analytics increasingly relies on open source software (OSS) libraries that extend scripted languages like python and R. Software documentation for these libraries is crucial for people across all experience levels, but documentation work raises many challenges, particularly in open source communities. In this collaboration between ethnographers and data scientists, we discuss the types, roles, practices, and motivations around documentation in data analytics OSS libraries.

Next Steps for the Colorado Risk-Limiting Audit (CORLA) Program

Abstract: Colorado conducted risk-limiting tabulation audits (RLAs) across the state in 2017, including both ballot-level comparison audits and ballot-polling audits. Those audits only covered contests restricted to a single county; methods to efficiently audit contests that cross county boundaries and combine ballot polling and ballot-level comparisons have not been available.

Colorado’s current audit software (RLATool) needs to be improved to audit these contests that cross county lines and to audit small contests efficiently.